Millennium, a hedge fund managing more than $92 billion in assets, is working with Anthropic to build a “digital risk analyst” — an AI teammate meant to help the firm’s human risk managers evaluate exposure across asset classes, the companies said August 6.
The tool, built on Claude, is designed to reason through risk positions, explain day-to-day changes in exposure, and recall context over time rather than starting each analysis from scratch. It runs in a sandboxed environment and produces recommendations that Millennium’s risk team must review and approve before acting on them, according to Anthropic.
“AI can help set a new bar for what our people can achieve, while keeping human judgment at the center,” said Millennium Chief Investment Officer Vlad Torgovnik.
Extending an existing deployment
Anthropic says Claude and Claude Code are already used across Millennium’s trading desks, engineering teams, and other core business functions, spanning more than 340 investment teams at the firm. The digital risk analyst extends that existing use into a more specialized function: assessing exposure for a hedge fund that trades across multiple strategies and finance markets.
Belinda Neal and Peter Nolan, who lead Anthropic’s financial-services and asset-management business respectively, framed the project as part of a broader push to make AI trustworthy in “complex, demanding environments” where mistakes carry a high cost.
The deal adds to a string of enterprise pushes by Anthropic this year, following moves such as Cognizant reaching the company’s top partner tier after training 30,000 staff on Claude, part of Anthropic’s tiered partner program.
Neither company disclosed a timeline for when the risk analyst will be deployed across Millennium’s operations, or whether Anthropic plans to offer a similar tool to other financial firms.